Cory Doctorow: IP
September 7, 2020 Commentary, Cory Doctorow
Photo by Paula Mariel Salischiker
You’ve probably heard of “open source software.” If you pay attention to the politics of this stuff, you might have heard of “free software” and even know a little about the ethical debate underpinning the war of words between these two labels. I’ve been involved since the last century, but even I never really understood what’s going on in the background until recently.
I was looking up the history of the first free software license, and I had a revelation that transformed the way I think about it, which is significant, since, to a first approximation, this stuff is all I think about.
Some background: “free software” had its origins with AI researcher-turned-activist Richard Stallman who started his GNU project in 1983, leading to the creation of the first “GNU General Public License” (GPL). This is a copyright license for computer programmers who want to share their work. If you release a program’s underlying source code under the GPL, anyone else is free to:
- Run your program
- Study your code
- Improve your code
- Share their improved code with others (provided that the same license is applied to that new code).
The GPL – a copyright license for software – arrived just as copyright for software itself arrived. Prior to 1983, software was generally viewed as a “functional work” and thus ineligible for copyright, but 1983’s Apple v. Franklin suit, combined with some 1980 amendments to the Copyright Act, established that software could be copyrighted.
By choosing to apply the GPL to their copyrighted code, programmers could “lock it open” and ensure not only that others could build on it, but that no one could take their work and close it off again.
Today, the act of licensing your code as “free software” is considered a mark of forbearance, a software author’s promise not to seek legal sanctions against those who create new programs that read and write their data, nor against those who “fork off” competing versions of their program.
To make your software “free” is to promise that its users will have freedom – as the saying goes, code is “free as in speech” even if it’s not necessarily “free as in beer.”
In 1998, free software underwent a reinvention, as a more business-oriented group of technical people and entrepreneurs decided that “free software” was alienating to the commercial sector, both because it stressed freedom and because it implied that the code would be free-of-charge and thus not directly commercializable.
Those people coined the term “open source,” to stress the instrumental benefits of allowing others to scrutinize code: these scrutineers could root out defects and prevent the nefarious subversion of code to spy on or steal data from its users (there’s a saying for this, too: “with enough eyeballs, all bugs are shallow”).
But let’s go back to 1983 and the birth of free software. When free software was born:
- software copyright was new and not widespread; it was not clear which activities were prohibited;
- software patents were almost unheard of and very hard to secure;
- terms of service (“shrinkwrap licenses” and other contracts of adhesion that you “agreed” to simply by running code, opening a box, or taking some other mundane, unsuspecting step) were not considered enforceable;
- there was no law against bypassing copy-controls or DRM;
- exotic “para-copyrights” – like the one that Oracle is currently trying to establish by suing Google over using an API – certainly didn’t exist.
In other words, when free software was born, if you wanted to clone, interoperate with, improve, or modify any existing software, all you needed to do was reverse-engineer that software and go to town.
The GPL wasn’t permission to make something new that was compatible with something that already existed. People who knew how to make software already behaved as though that was something they were permitted to do, irrespective of the feelings of the company or individual whose code they were interacting with.
The GPL was a request for companies and software authors to make this process easier and less tedious by handing out source code, so that would-be interoperators could skip the wasteful, time-consuming reverse-engineering step and go right to the good stuff – making a new thing that improved upon (and/or competed with) whatever was already made.
The GPL was icing on the cake. Applying the GPL to your code didn’t signal that you’d forswear legal vengeance upon those who wanted to make something compatible with your thing. You had no right to that vengeance! Applying the GPL to your code signalled that you wanted to collaborate with interoperators, rather than impotently shake your fist at them from the sidelines as they went right ahead and interoperated with your code against your wishes.
Some 40 years later, the world is a very different place. Between software copyrights, anti-circumvention rules, software patents, enforceable terms of service, trade secrecy, non-compete agreements, and the pending (at the time of this writing) Oracle/Google dispute over API copyrights, any attempt to interoperate with an existing product service without permission from its corporate master is a legal suicide mission, an invitation to almost unlimited civil – and even criminal! – litigation. That is to say: if you dare to modify, improve, or replace an existing, dominant software-based product or service, you risk bankruptcy and a long prison sentence.
Forty years ago, we had cake and asked for icing on top of it. Today, all we have left is the icing, and we’ve forgotten that the cake was ever there. If code isn’t licensed as “free,” you’d best leave it alone.
What is “interoperability,” anyway?
The term is nerdy, technical, obscure. It’s closely related to the slightly more familiar “compatability,” but the two aren’t quite equivalent.
In a technical sense, “interoperability” describes two products or services that can somehow work together with one another. From opening your Word documents in Google Docs, to using third-party ink cartridges in your printer to replacing your watch band, to changing the stereo that came with your car, interoperability is a broad, universal, essential characteristic of all of our technology.
Interoperability is the default state of the world. Anyone’s charcoal will burn in your barbecue, just as anyone’s gas will make your car go. Any manufacturer can make a lightbulb that fits in your light-socket and any shoes can be worn with any socks.
Some of this is down to standardization: manufacturers, academics, regulators, and interested parties gather in “standards development organizations” to make this process simpler, describing the canonical direction and spacing of a lightbulb-thread, or the syntax of an HTTP request, or the fittings on the underside of your toilet.
This certainly makes interoperability smoother! Standards for paper, from weight (“grams per square meter”/GSM) to size (letter/legal/tabloid; A1, A2, A3, A4, etc.) make it possible for you to reliably buy paper that will work with your printer, without requiring additional trimming or other modifications.
A failure to standardize can make life hard for everyone. Early Australian rail barons laid their tracks in several gauges, leading to the “multi-gauge muddle” of a rail system where some cars and engines could not run on some of the tracks.
These barriers to interoperability aren’t insurmountable. If your paper doesn’t fit your envelope, you can fold it; if it doesn’t fit your printer, you can trim it. If the rail gauge doesn’t match your rolling stock, you can modify the undercarriages to allow for multi-gauge operation (a difficult operation to be sure, never implemented despite hundreds of proposals) or you can tear up some of the track and lay new ones (as Australia has done and promises to do more of).
Interoperability lowers “switching costs” – the cost of leaving behind whatever you’re using now in favor of something you think will suit you better. When my grandparents emigrated to Canada from the Soviet Union on a displaced persons ship, they incurred a high switching cost: for more than a decade, they had no contact with their family in Leningrad except through unreliable, slow word-of-mouth with the rare person who got a visa to travel there.
Contrast this with my move from the UK to Los Angeles in 2015: we are in routine contact with my in-laws in London and Wales, as well as my family in Toronto. My laptop and books came with me, as did our other personal effects. We left most of our appliances behind because they ran on a different voltage, but there were a few things we loved that we brought with and either changed the plugs on or connected to our house’s electrical outlets via transformer or adapters.
Companies like high switching costs. For a would-be monopolist, the best product is one that’s seductively easy to start using and incredibly hard to get rid of. Think of Purdue Pharma’s gleeful internal memos – revealed in leaks and court cases – about the ease with which their “customers” were getting started on opioids and how hard it was for those same people to switch away.
Addiction isn’t the only way to raise switching costs. Facebook makes it incredibly easy to get started, historically going so far as to tricking you into giving it access to your electronic contacts list to enmesh you in a network of others who’ve already signed up for the service. Once you’re on Facebook, it’s very easy to bring in articles from the public web, and to link to your friends’ updates on rival networks. You can start by just using Facebook to follow the friends you have there, but over time, the system nudges you toward using Facebook as your primary means of reading the news and even following what your friends are saying on non-Facebook networks.
But when you want to leave Facebook, there’s no easy way to do so. You can’t go to a Facebook rival and follow what your friends post to Facebook from there. You certainly can’t reply to what your Facebook friends post using a rival service.
Interoperability – the thing Facebook uses to slurp stuff in from the open web – is the key to self-determination. Leaving Facebook in the 21st century is like my grandmother leaving the USSR in the 40s: you can go, but your friends and loved ones are all held hostage behind Zuckerberg’s Iron Curtain, so leaving Facebook means leaving your communities, your relationships. That’s not as hard as kicking opioids, but it’s not easy either. And your presence on Facebook is the reason someone else can’t go.
Here’s the thing: everyone wants to minimize risk, from employers to workers, from Big Tech to its users. You want to use Google in ways that make your life better, and you don’t want Google to be able to arbitrarily change or remove the services it provides (ask me how bitter I am about Google nuking Reader, its RSS product!).
Google wants to ensure that you won’t leave the company or its products and services. It could improve its retention by making you so delighted with its offerings that you never consider leaving – but a surer, cheaper way is to interweave its products and services with your life: making sure that your kid can’t go to a public school without creating a Google account; embedding Google search in your mobile OS; releasing web- and app-development frameworks for third parties that quietly harvest the data of their users and send them to Google.
The more freedom you have to leave Google, the bigger a risk you present to Google. The more Google can lock you in, the lower the risk of your departure from the service – and the higher the risk that Google will cease to keep your business by making good products, and instead rely on retaining you because you can’t leave (or because leaving comes at a very high price).
Interoperability improves self-determination by safeguarding your ability change the your current situation by incremental steps: if you like your phone and the apps you have, but want an app that’s banned in its default app store, interoperability comes to the rescue, allowing you to add a second app store to your phone’s list of approved software sources. You get to keep your phone, keep your apps, keep all the data on your phone, and you get to install that unauthorized app.
Without interoperability, your choice is “take it or leave it”: if the app store blocks an app you want, the price of getting that app is throwing away your phone, all its apps, and some or all of the data you’ve painstakingly input into your phone. That unauthorized app needs to be pretty darned good before anyone would pay such a high price for it.
Writ large, interoperability encompasses things like democracy: when someone says they like their city but not its bylaws, we don’t tell them that the law is the law and the home comes with these bylaws in a package. Instead, we set out processes for amending or repealing laws that chafe the people they govern. And, if you fail in your bid to reform your city’s laws, you can move to another city without having to surrender the possessions in your home or your social relations with your old neighbors. Interoperability lets you replace the laws and keep your house, or replace your house and find new laws.
This whole line of thought started with a reflection of the history of the free software movement: the largely forgotten time in which the default condition of software was freedom. In the absence of copyright, patent, anti-circumvention, terms of service, noncompetes, confidentiality, and other commonplaces of today’s software marketplace, anyone who could figure out how to reverse-engineer a program could improve it, replace some or all of it, read or write its files, compete with it, or sideline it.
Today, this is no longer the case. In fact, today’s software marketplace is so unlike this previous “cake-and-icing” world – where the default was software freedom (cake) and the free software movement began its audacious demand for freely reusable source code as a means of making software freedom as frictionless as possible (icing) – that it’s virtually impossible to imagine such an environment.
The thicket of anti-interoperability rules that has sprung up around interoperability has a catch-all name: “intellectual property.”
Now, free software advocates – and free culture advocates – hate the term “intellectual property.” The argument against IP rails against its imprecision and its rhetorical dishonesty.
Prior to the rise of the “intellectual property” as an umbrella term, the different legal regimes it refers to were customarily referred to by their individual names. When you were talking about patents, you said “patents,” and when you were talking about copyrights, you said “copyrights.” Bunching together copyrights and trademarks and patents and other rules wasn’t particularly useful, since these are all very different legal regimes. On those rare instances in which all of these laws were grouped together, the usual term for them was “creator’s monopolies” or “author’s monopolies.”
The anti-IP argument leans into the differences between the underlying rationale for each of these rules:
- US copyrights exist to “promote the useful arts and sciences” (as set out in the US Constitution); that is, to provide an incentive to the creation of new works of art: copyright should offer enough protection to create these incentives, but no more. Copyright does not extend to “ideas” and only protects “expressions of ideas”;
- Patents exist as incentive for inventors to reveal the workings of their inventions; to receive a patent, you must provide the patent office with a functional description of your invention, which is then published. Even though others may not copy your invention during the patent period, they can study your patent filings and use them to figure out how to do the same thing in different ways, or how to make an interoperable add-on to your invention;
- Trademarks exist as consumer protection: trademarks empower manufacturers to punish rivals who misleadingly market competing products or services that are like to cause confusion among their customers. It’s not about giving Coca-Cola the exclusive right to use the work “Coke” – it’s about deputizing Coca-Cola to punish crooks who trick Coke drinkers into buying knockoffs. Coke’s trademark rights don’t cover non-deceptive, non-confusing uses of its marks, even if these users harm Coca-Cola, because these do not harm Coke drinkers.
Seen in this light, “intellectual property” is an incoherent category: when you assert that your work has “intellectual property” protection, do you mean that you can sue rivals to protect your customers from deception; or that the government will block rivals if you disclose the inner workings of your machines; or that you have been given just enough (but no more) incentive to publish your expressions of your ideas, with the understanding that the ideas themselves are fair game?
When you look at how “IP” is used by firms, a very precise – albeit colloquial – meaning emerges:
“IP is any law that I can invoke that allows me to control the conduct of my competitors, critics, and customers.”
That is, in a world of uncertainty, where other people’s unpredictability can erode your profits, mire you in scandal, or even tank your business, “IP” is a means of forcing other people to arrange their affairs to suit your needs, even if that undermines their own needs.
There are some ways in which this is absolutely undeniable. Take DRM, “Digital Rights Management.” These are the digital locks in our devices that prevent us from using them in ways that the manufacturer dislikes. Your printer uses DRM to force you to buy ink that the manufacturer has approved; your phone uses DRM to force you to buy apps that the manufacturer has approved. Ventilators from Medtronic and tractors from John Deere use DRM to force you to get them repaired by the manufacturer – and to scrap them when the manufacturer decides it’s time for you to buy a new one.
Copyright laws – that is, “IP laws” – ban tampering with DRM, making it a serious, jailable felony to provide others with tools to bypass DRM. From Section 1201 of the US Digital Millennium Copyright Act to Canada’s Bill C-32 to Article 6 of the EU Copyright Directive, countries around the world have imposed indiscriminate bans on breaking DRM.
These are all copyright laws, but, tellingly, the ban on breaking DRM is not limited to copyright infringement. Bypassing DRM to get your printer to accept third-party ink is not a copyright violation: you’re not reproducing its code, nor are you duplicating the traces etched into its chips. But even though you’re not breaking copyright when you jailbreak your phone, you’re still breaking copyright law. The law bans legal conduct, if you have to break DRM to engage in it. This isn’t copyright protection – it’s felony contempt of business-model.
It’s not just DRM. Take “Goldman Sans,” a free font released by the finance giant and global supervillain Goldman Sachs. Goldman Sans is a copyrighted work, and it comes with a copyright license that you “agree” to when you download the font. Among the license terms for Goldman Sans is a non-disparagement clause – that is, a clause that prohibits you from criticizing Goldman Sachs. Goldman Sachs doesn’t need copyright law to prevent people from copying its font. It gives the font away for free. Goldman Sachs needs copyright law so it can boss people around – so it can tell them what they may (and may not) say.
The risks to free expression and self determination have always been latent in copyright, patent, and trademark, and these laws have historically been designed to minimize those risks. Each one has its own “escape valve” that, theoretically, stops “IP owners” from using their rights to take away your rights.
Copyright has “fair use” (“fair dealing” in most non-US English-speaking countries), which allows for many kinds of copying, adapting, displaying, and even selling of others’ copyrighted expressions, provided these activities promote a free and robust discourse by transforming, commenting on, or analyzing the copyrighted work. Fair use doesn’t depend on a copyright holder’s permission – you can make fair uses even (especially!) if the rights-holder doesn’t want you to.
Patent has its own escape valve: publication. To receive a patent, you must disclose how your invention works and those disclosures are on display from the start, where anyone can study them and use them as inspiration for their own inventions. Patents allow you to punish people who duplicate your invention, but they also require that you tell people exactly what steps they must take to effect such a duplication, and also provides a roadmap for replicating your invention’s functions without violating your patent.
Trademark has two important escape valves: first, trademark holders are limited to enforcing their marks against rivals who use them in deceptive ways likely to cause public confusion; second, trademark is subject to the “nominative defense” – it’s not a violation of a trademark to use that mark to describe the goods or services it’s associated with. You can put a sign in your shop window reading, “We fix iPhones,” or “Cheap ink for HP printers” or “Our cola tastes better than Coke!” and there is nothing the trademark holder can do about it.
These escape valves have been a lot less durable than we might hope. It turns out that much of their efficacy depends on there being robust competition in the marketplace, so that when one company tries to narrow (say) fair use in court, other companies that depend on fair use spring up to defend it. Through the past four decades of massive consolidation in every industry, a consensus has emerged among the shareholder and managerial classes that these escape valves are defects in otherwise excellent laws, and they have set to work creating legal precedents, new laws, and new legal tactics to jam these valves shut.
This is how we went from having software freedom cake to just having the icing: new copyright laws (like the ones that ban breaking DRM); new copyright precedents (like the one Oracle is presently seeking in its lawsuit against Google); and new tactics for combining copyrights, patents, trademarks, DRM, trade secrets, and other IP so that what trademark permits, copyright prohibits, and what copyright permits, patent blocks, and so on, until all the certainty has been moved onto the manufacturer’s side of the deal, and all the risk has been moved onto yours.
Recall that the term of art that preceded “IP” was “author’s monopolies.” The tale of how the latter was replaced with the former has many variations, but everyone agrees that its no fun to be called a monopolist.
If you think you don’t have enough copyright, it’s hard to go to Congress or Parliament and demand an expansion of your regulatory monopoly – far more pleasant is demanding help in defending your “property.”
There is a strict sense in which copyrights and (especially) patents are monopolies. When you write down a collection of words – this essay, say – a new copyright is born. It gives you the exclusive right to reproduce, adapt, display, or sell the essay. Your rivals can compete with you by writing different essays that they hope to sell to the same magazines to tempt the same readers with, but you and you alone can sell the essay you wrote. In that sense, it’s obviously a monopoly – a market for a product with only one seller, which is the purest form of monopoly there is.
But the word monopoly has a different definition in competition law: a monopolist isn’t merely someone who controls 100% of the supply of a good or service. A monopolist is anyone who has “market power” – that is, the power to set prices. After HP sells you the printer, it can charge whatever it wants for ink. You could junk your printer and buy another, but HP’s rivals also charge absurd sums for ink and also use DRM and IP to punish competitors who introduce cheap ink into the market.
The people who bristled at having copyrights called “monopolies” had a point: it’s a rare author who has “market power.” Authors – even big name ones – have limited power to convince their publishers to pay over the odds (there are a few superstar exceptions, but they’re as rare as meteor-strikes; overwhelmingly, copyrights do not confer market power to creators).
However, there are actual, market-power monopolies in the entertainment industry: the single movie-theater chain that controls the vast majority of cinema screens (AMC, which may be bankrupt or a part of Amazon by the time you read this); the three record labels, four movie studios, and five publishers (maybe four by the time you read this); the single national brick-and-mortar bookstore chain and the single global online bookseller (which also effectively owns the audiobook market).
How did these monopolies emerge? Well, in some ways, they came from the same place that all the other monopolies – in energy, eyewear, finance, automotive, aerospace, accounting, civil engineering, logistics, etc. – came from: lax antitrust enforcement. For 40 years, we’ve let companies grow by buying their small competitors before those competitors could become threats, to merge with their major competitors and cease competing, and to corner vertical markets so they could squeeze labor, suppliers, and customers.
Entertainment monopolies are special, because they aggregate these “author’s monopolies” in vast quantities. These monopolies are durable in ways that mere market power is not. If you control one third of all the music that might be sampled by other musicians, you have a bottleneck that can’t be evaded through cunning or creativity. Any attempt to break your monopoly is a copyright infringement, and doing it at commercial scale is a criminal copyright infringement.
That is to say, banking and aerospace monopolies can get sued for being anticompetitive – but entertainment monopolies can sue you for being pro-competitive. The result is a monopoly that controls access to distribution channels and audiences – that can invoke the power of the state to fine or even im prison people who seek to challenge that monopoly.
Even where tech is challenging these monopolies, it is doing so in order to create more monopolies. Kindle Unlimited presents a real challenge to traditional genre publishing, and every Kindle Unlimited book is released with DRM that locks it to Amazon’s platform. Any attempt to liberate Kindle Unlimited books so they can be read with a rival’s device (or with a device designed to stop Amazon from spying on you while you read) involves breaking the DRM, and trafficking in tools to break DRM is a felony under Section 1201 of the Digital Millennium Copyright Act of 1998. Kindle Unlimited books aren’t available in libraries, and giving librarians a tool to remove the DRM to add Kindle Unlimited books to their collections is a jailable offense.
As Amazon conquers an ever-larger proportion of genre readers, it permanently locks those readers into its platform, meaning that any author who wants to access those readers will have to do so on Amazon’s terms, turning over the power of their “author’s monopoly” to be used in Amazon’s “market power monopoly” arsenal. Give authors more copyright – a stronger monopoly – and Amazon will seize that too, as a condition of reaching the audience Amazon has imprisoned in its walled garden.
“Author’s monopolies” are not “market power monopolies,” but if you can aggregate enough “author’s monopolies” in one place, you can turn it into a market power monopoly that is backed by the power of the courts and the prisons and that accumulates more author’s monopolies every time someone enters your captive marketplace.
You may have heard Netscape founder Marc Andreessen’s famous phrase: “Software is eating the world.” It’s not quite true. Software has eaten the world.
If there was any doubt, the pandemic erased it. Locked down, viewing the world through our screens, there is no longer any distinction between human rights and digital rights. There is no software freedom. There’s only freedom.
Software is expanding relentlessly into every device and system, from ventilators to tractors, from toothbrushes to sex-toys, from refrigerator filters to money itself. Wherever we find software, we find “IP” – that is, software deployed to control the conduct of competitors, critics, and customers, in such a way that overlapping systems of laws can be invoked to punish anyone who bypasses that software.
The combination of software and IP in every device is a sea-change for the organization of our society. As firms have become increasingly concentrated – as monopolies have emerged in every sector – they have also figured out how to infuse their products with just enough software that they can invoke IP to control their competitors, critics, and customers. These are market power monopolies backstopped by creators’ monopolies, which create IP rights that supercharge their market power.
All of this is just a curtain-raiser. Software isn’t just a way to put IP into otherwise inert objects. It’s also a way to automate them, to make them into unblinking, ever-vigilant enforcers for the manufacturer/monopolist’s interests. They can detect and interdict any attempt at unauthorized interoperability, and call the appropriate authorities to punish the offenders.
This is a level of control beyond the wildest dreams of history’s most sociopathic monopolist. Consider the coal-boss who controlled his workers by moving them into company housing in a company town where they were paid in company scrip that could only be spent at the company store. This coal boss moves titanic amounts of risk off his balance sheet and onto his workers’. Not only are they incapable of leaving for a better job without paying their debts, but they also are paid in non-interoperable, proprietary money that only works at the company store, where prices can be adjusted at will to ensure that the workers’ debts are never paid.
Even that coal-boss, a God to his workers, was not all-powerful. A coal worker could buy corn at the company store and trade it for real US greenbacks at the local moonshiner’s shed, converting non-interoperable scrip to interoperable dollars at a loss, through the intermediary exchange medium of corn.
Think of a worker paid in company scrip today – the digitally enabled smart goods they buy at their company store can be locked to their accounts the way a Kindle book or an iPhone app is locked to your personal device. The unblinking eye of the software enforcement system is always watching, ready to discipline you for your lack of consideration for its shareholders’ bottom line.
Once we had cake. Today we have icing. At this rate, the icing will be gone before long.
There are no digital rights, only human rights.
There is no software freedom, only human freedom.
Cory Doctorow is the author of Walkaway, Little Brother, and Information Doesn’t Want to Be Free (among many others); he is the co-owner of Boing Boing, a special consultant to the Electronic Frontier Foundation, a visiting professor of Computer Science at the Open University and an MIT Media Lab Research Affiliate.